Obviously, both comets, 26P/Grigg-Skjellerup and 73P/Schwassmann-Wachmann 3, have a very small size as compared to other periodic and non-periodic comets (see Meech 1998). In fact, they are among the smallest nuclei of short-period comets and only 46P/Wirtanen (Boehnhardt et al. 1996) and 45P/Honda-Mrkos-Padajowskaja (Lamy et al. 1996) seem to be even smaller objects.
Our equivalent radius estimation of 1.5 km for 26P/GS is larger than the one determined by radar observations (Kamoun 1983, Kamoun et al. 1996), but it is smaller than the values deduced from other ground-based CCD photometry by Birkle & Boehnhardt (1992; 2-3 km) and by Meech (1998; 2.9 km). Unfortunately, GIOTTO measurements could not be used for the size estimation of this nucleus. Considering the uncertainty of the radii from the CCD photometry, we conclude that this nucleus is either of non-spherical shape with a maximum axis ratio (small-to-large) of 0.5 and/or it has a rather variable surface albedo (the large crust coverage of the surface - see discussion below - makes this scenario less likely). The latter axis ratio is in disagreement with the one determined by Sitarski (1992; 0.7) from an analysis of the non-gravitational forces on the cometary orbit. The larger axis ratio (0.9) determined from our lightcurve of the comet could be due to a different viewing geometry of the effective nucleus cross section. In summary, although the exact value of the axis ratio may still be debatable, it is very likely that Comet 26P/GS has a non-spherical nucleus. The rotation of the nucleus seems to be longer than 12 h, a conclusion supported by our photometry of the nucleus lightcurve. The observed coma fans (Birkle & Boehnhardt 1992, Fulle et al. 1993) are at least in qualitative agreement with a fast rotation period of the order of one day or less. Using our value for the nucleus size (1.5 km) and model calculations for the gas release from water ices (Huebner 1992), one can conclude that the water production rate observed during the GIOTTO encounter period (6-7 molecules/s: see Neubauer et al. 1993 and Jockers et al. 1993) required only about 5 percent of the surface being active. At least some - if not a major amount - of this activity must have been concentrated in an active region which produced the coma fan during the last two apparitions (Birkle & Boehnhardt 1992, Fulle et al. 1994). Therefore, a large crustification of the nucleus of 26P/GS has to be assumed. The nucleus crust and the different illumination geometry of the comet by the Sun may also explain the absence of a weak coma in our September 1993 (post-perihelion) observations. In conclusion, Comet 26P/GS has a very small, non-spherical and highly crusted nucleus which may be typical for a very evolved and old periodic comet. Nevertheless, the measured slightly bluish V-R colour does not comply with our general picture of red colour gradients for evolved cometary nuclei. The nucleus fragments which are hypothesized from OPE observations onboard GIOTTO (McBride et al. 1997) must have been beyond the detection limit of our combined data of 9.5h total integration (the detection limit of 27 mag corresponds to about 200m equivalent radius) or already outside of the field of view of our detector (6.9 arcmin corresponding to 860000 km projected distance at the distance of the comet).
Our 1.1 km radius for 73P/SW3 gives only an upper limit for the size of this comet (because of the light contamination from the weak coma and tail). It is, however, in surprisingly good agreement with the radius value which is obtained from the visual brightness estimations of the central coma condensation (considered to represent the nucleus) made by Baldet during the close approach of this comet to Earth in 1930 (Baldet 1930a,b). Sekanina et al. (1998) calculated a radius of about 1 km from this rather coarse visual photometry of the comet. Obviously, 73P/SW3 had a very small nucleus already before it broke apart in 1995 (Boehnhardt & Käufl 1995; Sekanina et al. 1998 and references therein). If the sub-nuclei can be detected after the dust and gas emission of the split fragments has ceased, it may be possible to measure and compare directly - for the first time ever - the size of the parent body with those of its fragments. Shortly before break-up in 1995 the comet was hyperactive (Sekanina et al. 1998) as can be concluded from the observed OH production rates (Crovisier et al. 1996). The effective radius to support this activity is more than 2 km (Sekanina et al. 1998) which is at least a factor 2 larger than our upper limit. The former authors suggest that evaporating icy grains may have contributed a significant amount of OH in the outburst phase before the break-up of the comet. Considering the fact that during the past apparition the lightcurve of the comet followed the normal development until the outburst in August/September 1995 and assuming a proportional down-scaled water production rate as compared with the outburst level (which was 22.2 molecules/s; Crovisier et al. 1996), one arrives at a nucleus radius of 0.2 km which could produce the typical perihelion activity if 100 percent active. This would mean a crustification of the surface of around 96 percent or in other words only about 4 percent (minimum) of the surface area were needed to be active in order to support the typical perihelion activity of this comet. The nature of the variability in our comet photometry of 73P/SW3 is not known (secular brightness increase or effects from the nucleus rotation). At the time of our observations the comet had already started the pre-perihelion activity. One month later, the comet appeared about 2 mag brighter (own unpublished observations of 29 January 1995). However, in mid 1994 the comet could not be detected with the 3.5m Calar Alto telescope (again own unpublished observations). Hence, the small coma diameter in December 1994 may indicate that the activity onset was not very strong and/or that we observed the comet in a very early state of gas and dust production, i.e. during activity onset. In summary, 73P/SW3 is another example of a small cometary nucleus, but it may be different in its crustification and gas and dust emission since it developed from a (likely) very crusted into an hyperactive object. Its V-R colour is very similar to that of 26P/GS and does not show any reddening effect (as expected for old and evolved objects). Finally, our conclusions on the possible crustification of the nucleus can be jeopartized by the two main assumptions used for our calculations, i.e. that the effective nucleus radius is close to our measured upper limit and that the water production rate near perihelion scales according to the visual brightness of the comet. If this is not the case, the nucleus of P/SW3 would be more active and with less crust coverage.
© European Southern Observatory (ESO) 1999
Online publication: December 16, 1998